Digital Evaluator - An App for Trial Gardens - OSU WebGarden

Digital Evaluator - An App for Trial Gardens
Tim Rhodus, Professor and
Elaine Eberlin, Systems Specialist
Department of Horticulture & Crop Science, The Ohio State University
[email protected] and [email protected]
Online at: webgarden.osu.edu/buckeye2015.pdf
This article describes the Ohio State University Cultivar Trials and how an app
was created to assist in organizing and managing the evaluation and
communication program for the garden.
Initially conceived by Dr. Claudio Pasian in 2002, the Ohio State University Cultivar Trials are
conducted each year in order to evaluate ornamental bedding and container plants. The
objective of these trials is to observe the performance of new and recently introduced cultivated
plant varieties under environmental conditions typical of central Ohio.
According to Dr. Pasian, “Over the years, we have grown and put ourselves on the map of
annuals trials in the U.S. Growers from around the country and abroad visit the annuals trial
every July during the Cultivate event in Columbus. In August, with the help of the Ohio
Nursery and Landscape Association (ONLA) and AmericanHort, we also have an open house
geared to landscapers.”
The OSU Cultivar Trials contain four categories of plants: field (ground beds), containers in the
sun, containers in the shade, and hanging baskets. During the growing season, all plants are
evaluated four times by Master Gardener and Chadwick Arboretum volunteers. These
consumer evaluators are asked to rate each plant from 1 – 5 on how well they “like” the
plant. Additionally, the Trial Director and Trial Manager do a more in-depth “grower”
evaluation. Results and photos of the 2014 trials can be found at: hvp.osu.edu/trials
Moving Towards a Digital Solution
Traditionally, each evaluator was given a clipboard and 20 pages of plant names that needed to
be evaluated on four different occasions. Volunteers and Trial Staff picked up their clipboard
and walked up and down the beds writing their scores in the appropriate boxes. The same
clipboards were picked up a month later and used to enter more results. At the end of the
season, a student worker was recruited to enter all the scores in a spreadsheet that was then
used to generate overall average ratings. Considering that over 500 entries were evaluated each
year by approximately 20 volunteers on four different occasions, this approach to data
gathering and information management was not only inefficient, time consuming and expensive
but also prone to error (Figure 1).
2 Figure 1: One of twenty pages used to evaluate plants
Dr. Rhodus first approached the Cultivar Trial organizers, Dr. Pasian and Lindsay Pangborn, in
May 2014 with his idea for developing a mobile app that could be used to not only record
evaluation data but also transmit it to a server. That way, the process of recording, transmitting
and sharing evaluation scores could be done without the need for clipboards or student workers.
In addition, the new process would help the evaluators do their work quicker, the data could be
viewed shortly after each evaluator finished their work and the app would be capable of taking
pictures of the plant being evaluated.
Deciding What to Include
During follow-up meetings, Dr. Rhodus learned about the list of plants in the trial and what
information needed to be collected during each of the consumer and grower evaluations.
However, the biggest challenge was learning where 563 different plants were located in 44
different beds both in front of and behind Howlett Hall.
For Claudio and Lindsay, this had become second nature. However, as newcomers, we needed
to see a map that was both accurate and visual, not a hand-drawn representation. So,
we used Google Maps, satellite view and zoomed into the appropriate areas north and south of
Howlett Hall and took screen captures. Given that an evaluator could start with any category of
plant or with any specific bed, the maps had to display all 44 planting beds and serve as the
primary linking mechanism to the subset of plants in the database that were to be evaluated at a
given location. This was done through the addition of clickable icons on a map, each
represented by a colored button indicating category abbreviation and bed number. For
3 example, F-1 was a white button for Field category bed #1 and C-8 was a red button for
Container category bed #8 (Figure 2).
Figure 2: Locations of planting beds Sun Screen Protection
A significant factor affecting usability but wasn’t part of app design, involved using iPads or
iPhones outside in bright sunlight. It was obvious from the beginning that being able to view a
mobile screen or enter data in bright conditions was going to be a problem and if we couldn’t
4 find a solution, there would be no going forward. Success!!! After researching various screen
filters, we decided on NuShield’s product called DayVue™. This film is used widely by pilots,
boat owners and even the military. You can select the specific make and model of your mobile
device at the following: nushield.com
The Software
The development goal was to create a self-contained app that would work on iPhone or iPad
and not require any wifi connection. We wanted a design where evaluators could walk around
holding their mobile device and easily enter evaluation data. Lastly, there was less than four
weeks to develop the app. In order to meet these goals, we decided to use FileMaker Pro
software to create a database that would serve as the foundation for the app. We've used this
software for over 20 years of web development projects and have found it to be very powerful
and easy to use. FileMaker also offers web server and mobile device products. These three
programs are fully compatible with each other and were all used in the project.
We also learned that the planting information
was contained in spreadsheet files. This was
not a problem, as a spreadsheet file can be
easily imported into a FileMaker database.
The final part of the project was to download
and open the finished database in FileMaker
Go. This free mobile app enables anyone to
design a fully functioning mobile app that
supports all the fetures of a database.
Primary Features of Digital
Evaluator
Our finished program was named, Digital
Evaluator. Building on our web design
experiences, the initial page of this app is
referred to as Home. This layout connects a
user with the basic administrative and
evaluative controls they need to participate in
the Cultivar Trials (Figure 3).
Evaluator ID - All evaluators (new and
returning) were trained in how to use the app.
As part of their training, they were each
assigned an ID number. This was selected on
Home page and linked to all scores recorded
and transferred by that evaluator.
Figure 3: Digital Evaluator app for iPhone 5 Category - The evaluator selects which category of plants they want to evaluate – Field (north
of Howlett Hall) or Basket, Container, Shade (south of Howlett Hall). Tapping the
appropriate button links to the detailed navigation map for selecting individual beds.
Progress Report - With 563 plants to evaluate, it's easy to miss some. This report shows the
evaluator which plants have not been evaluated and allows him/her to return to the appropriate
bed and enter a score.
Export Data – After completing the evaluation, those people using their own device need to
export the data. Step-by-step instructions are contained in a special layout. Evaluators who
borrowed a device simply returned it and the export was done by us.
Evaluation Layout
A single evaluation layout was developed for use with all plant categories. However, before
entering a score, it was critical that evaluators confirm the plant they are looking at is the same
as the plant in the database. To assist with this, the app layout was designed to mimic the look
of the plant label (Figure 4).
Figure 4: Digital Evaluator App for iPad 1 – Evaluation screen Having confirmed they are rating the correct plan, the evaluator tapped the appropriate spot on
the evaluation scale to enter their score. Custom notes could be entered in the text box. The
evaluator walked to the next plant in the bed and tapped the Next button on the app. After
completing a given bed of plants, the evaluator would tap one of the Category buttons to return
6 to the primary navigation map and select another bed. Many of the returning volunteers
indicated the app helped them rate the plants in less time than using the old pencil and paper
approach.
A grower version of the app was developed to provide Claudio and Lindsay with evaluation
options that were essential to them as grower professionals but too in-depth for the consumer
version. An added benefit of the grower version was the ability to take pictures while using the
app, for both the iPhone and iPad versions. With this approach, we didn’t experience the usual
problem of trying to match photos taken with a digital camera to an exact plant in the database.
Training
To introduce the app, we set up training sessions for all new and returning evaluators. A trifold training brochure was developed to describe the trials and depict the various features of the
app. In total, we trained 54 volunteers in all aspects of using the app and reviewing the
progress report. For many of the volunteers, this was their first experience using an iPad.
Some of the volunteers preferred to use their own iPhone or iPad device. For them, an
extended training session was held to assist these volunteers with downloading and installing
the Digital Evaluator App on their devices and making sure they were familiar with the app,
how to adjust settings, how to export their results and how to update the app.
Website Development
The website - 2014 Ohio State Cultivar Trials (hvp.osu.edu/trials) showcased the data being
collected throughout the season. Using the app, volunteer evaluations were collected over the
course of one week, on four different occassions. Data was transferred from each device and
emailed to Dr. Rhodus, who then checked for completeness and uploaded it to FileMaker
server. When the evaluation window was closed, the average score for each plant, across all
evaluators, was compiled and saved in the database. This way, the database server program
does not have to recalculate average scores for all 563 plants every time someone looks at the
website.
Another feature of our website is the ease with which one might search/sort results by:
Category, Crop, Series, Cultivar, Company or Rating (Figure 5). Filtering of results for any of
these attributes is immediate. Compare this to other trial gardens where results are provided as
a spreadsheet file in “pdf” or “doc” format that requires downloading and oening in a separate
program - not very immediate and definitely not very flexible.
For those who are more visual, pictures taken with the app are also served on the website. The
Gallery not only shows the plants, but provides all of the important identifying information,
too, including Category, Bed, Location, Crop/Series/Cultivar, Company, and Overall Rating
(Figure 6).
7 InVirtual Perspective Technology Team...
The OSU inVirtual Perspective Technology Team consists of Dr. Tim Rhodus, Professor; Bud
Witney, Systems Manager, and Elaine Eberlin, Systems Specialist. The team is responsible for
the design and maintenance of the systems, databases, and much of the content contained in the
numerous web sites positioned within Horticulture in Virtual Perspective, OSU PlantFacts, and
American Society for Horticultural Science. Online access to this document is available at:
hvp.osu.edu/buckeye2015.pdf
Figure 5: Filter evaluation results for Coleus
8 Figure 6: OSU Cultivar Trials Website